Search

Glioblastoma (GBM) is the most prevalent and lethal primary intrinsic brain cancer. The disease is essentially incurable, with glioblastomas characterized by resistance to both chemotherapy and radiotherapy, as well as by rapid tumor progression, all of which are mainly ascribed to glioma stem‑like cells (GSLCs). In the present study, an improved model that is more similar to clinical GBM was constructed. Twenty clinical glioma samples were collected to obtain primary low‑grade tumor cells. The cells were either maintained in serum‑free medium as primary glioma‑based cells (PGBCs) or cultured in the same medium with CHIR99021 as GSLCs. Then, the molecular and ultrastructural differences between the two cell groups were determined. Furthermore, the proliferation and migration of the GSLCs were examined and the potential mechanisms were investigated. Finally, temozolomide resistance in vitro and in the mouse model was assessed to study the properties of the induced GSLCs. The primary low‑grade tumor cells extracted from surgical samples were enriched with GSLC properties, with high expression levels of CD133 and Nestin in 100 nM CHIR99021. The GSLCs exhibited high proliferation and migration. Furthermore, the expression of the PI3K/AKT signaling pathway and that of related genes and proteins were significantly enhanced by CHIR99021. The animal study also revealed high levels of STAT3, mTOR, NF‑κB, and VEGF in the GSLC‑transplanted mice. CHIR99021 could stably enhance GSLC properties in patient‑derived glioma samples. It may provide a useful model for further study, helping to understand the pathogenesis of therapeutic resistance and to screen drug candidates.

Patient‑derived orthotopic glioma xenograft models are important platforms used for pre‑clinical research of glioma. In the present study, the diagnostic ability of magnetic resonance imaging (MRI) was examined with regard to the identification of biomarkers obtained from patient‑derived glioma xenografts and human tumors. Conventional MRI, diffusion weighted imaging and dynamic contrast‑enhanced (DCE)‑MRI were used to analyze seven pairs of high grade gliomas with their corresponding xenografts obtained from non‑obese diabetic‑severe‑combined immunodeficiency nude mice. Tumor samples were collected for transcriptome sequencing and histopathological staining, and differentially expressed genes were screened between the original tumors and the corresponding xenografts. Gene Ontology (GO) analysis was performed to predict the functions of these genes. In 6 cases of xenografts with diffuse growth, the degree of enhancement was significantly lower compared with the original tumors. Histopathological staining indicated that the microvascular area and microvascular diameter of the xenografts were significantly lower compared with the original tumors (P=0.009 and P=0.007, respectively). In one case, there was evidence of nodular tumor growth in the mouse. Both MRI and histopathological staining showed a clear demarcation between the transplanted tumors and the normal brain tissues. The relative apparent diffusion coefficient values of the 7 cases examined were significantly higher compared with the corresponding original tumors (P=0.001) and transfer coefficient values derived from DCE‑MRI of the tumor area was significantly lower compared with the original tumors (P=0.016). GO analysis indicated that the expression levels of extracellular matrix‑associated genes, angiogenesis‑associated genes and immune function‑associated genes in the original tumors were higher compared with the corresponding xenografts. In conclusion, the data demonstrated that the MRI features of patient‑derived xenograft glioma models in mice were different compared with those of the original patient tumors. Differential gene expression may underlie the differences noted in the MRI features between original tumors and corresponding xenografts. The results of the present study highlight the precautions that should be taken when extrapolating data from patient‑derived xenograft studies, and their applicability to humans.

The primary human acute myeloid leukemia (AML) cell injection xenograft mouse model is used to investigate multimodal therapies and drug screening on tumor growth. Since xenograft models using human cell lines to examine drug response are not correlated with the clinical outcomes observed in patients, a xenograft model using primary human cells has been used as a more appropriate model with which to minimize this problem. Although bone marrow (BM) cells from patients are often regarded as superior sources to establish xenograft models due to the high frequency of stem cell populations, there is a fatal drawback; only small volumes can be obtained and used for the generation of the leukemic xenograft model. Indeed, longevity of AML characteristics, as well as sufficient stem cells in the xenograft model, should be guaranteed to analyze the therapeutic response to a drug. Therefore, we examined whether leukapheresed peripheral blood (LPB) consists of reliable leukemic stem cells (LSCs) and ALDHdim‑expressing CD34+CD38- cells, and functions in grafting human AML with virulence compared to that of BM. LPB cells showed an advantage for the xenograft mouse model with AML cell homing, engraftment and a high human ALDHdim-expressing CD34+CD38- cell population, suggesting an alternative cell source to BM. Overall, this xenograft model using LPB offers the possibility of overcoming the small volume limitation of BM and prevents individual variation by using a single LPB sample. This result is noteworthy in identifying cell sources capable of generating a stable xenograft model.

Glioma stem cells (GSCs) are key in the progression and recurrence of glioblastoma. Inducing the differentiation of GSCs is an important therapeutic target for glioblastoma. Nuclear factor erythroid 2-related factor 2 (Nrf2) has been reported to be important in maintaining the stem cell status of GSCs; however, its association with differentiation has not been studied. Herein, we knocked down Nrf2 from GSCs to investigate the role of Nrf2 in the differentiation of GSCs. First, Nrf2 expression was observed at different stages of differentiation; then, Nrf2 was knocked down and the association of Nrf2 with differentiation degree was observed in vitro. Finally, GSCs were planted in nude mice to study the association of Nrf2 with differentiation in vivo. The expression of Nrf2 decreased with the differentiation process. Following Nrf2 knockdown, the proportion of sphere-like colonies decreased and the dendritic cells in spheres increased; the expression of Nrf2 significantly decreased while the expression of differentiation marker glial fibrillary acidic protein (GFAP) and βIII-tubulin increased both at the protein and the gene level. In the xenografts of nude mice, the differentiation of tumor cells was improved. These results suggest that Nrf2 is a key factor inhibiting the differentiation of GSCs, and knockdown of Nrf2 may promote the differentiation process, providing a therapy target for GSCs.

Side population (SP) cells are a small subpopulation of cells with enriched source of gastric tumor-initiating cells (TICs) with stem-like cell property that are characterized by high efflux ability of Hoechst 33342 dye, reflecting high expression of several subtypes of the ATP-binding cassette transporter family that is characteristic of stem cells. The present study is the first to discover and characterize SP cells within gastric cancer (GC) tumors. In this study, human GC cell lines (MKN45, KATOIII, MKN74, MKN28 and MKN1) were analyzed using flow cytometry for SP cell isolation, and all GC cell lines showed a distinct fraction of SP cells, ranging from 0.02+/-0.001 to 1.93+/-0.16%. Among these cell lines, MKN45 cultures possessed the highest percentage of SP cells. Using MKN45 cells, we demonstrated stem cell-like characteristics of SP cells of the cell lines as a possible subpopulation with enriched TICs, as indicated by ABC transporter gene expression (MDR1 and BCPR1), chemo-resistance and tumorigenicity in vivo. In addition, we report the first identification and isolation of SP cells from clinical GC tissues as well as human GC cell lines. These SP cells demonstrate higher tumorigenicity in vivo than does the overall cell population in the parent tissue. In conclusion, we demonstrate that solid tumor tissue such as human GC contains TICs, with the existence of heterogeneity and distinct hierarchy in malignancy, suggesting the future possibility of a novel therapeutic tool targeting TICs for overcoming this malignant disease.